Full-text search across 68501 articles. Combine with topic and date filters; results sorted by relevance.
A developer has launched Radio, a provider-agnostic communication layer that lets AI agents from different vendors and machines join a single shared conversation via a URL. Built as a core primitive of the team's Fractal…
Comma.ai published a blog post detailing machine learning bugs that degraded its self-driving model's training and inference, including a BF16 plan-output precision issue that limited speed representation to 0.125 m/s st…
Monid.ai launched a router for agent tools that gives AI agents access to more than 2,000 tools across 72+ providers through one base URL and one API key, with usage metered per call. The connector layer is declarative a…
OpenAI's agentic software factory uses a risk classifier to route pull requests it labels "low risk" directly to production without human review when specialist reviewers find no blockers. The system's routing decision, …
Security researchers disclosed CVE-2026-48710, dubbed BadHost, an authentication bypass in the Starlette ASGI framework caused by inconsistent parsing of malformed Host headers. The flaw, affecting Starlette versions 0.8…
A developer traced the network calls of an agentic coding assistant and found that a simple two-file PyQt edit consumed 3 LLM calls and roughly 760 KB of exchanged JSON, versus a single call and about 100 KB with Aider. …
A developer analyzed 35 days of Claude Code MCP connection logs covering 27,257 connections and found that OAuth token refreshes inside the connection window make sessions 5.8× slower (3,170 ms median vs 550 ms), with th…
A developer building agent UIs argues that teams should decide where the agent loop runs before designing the chat interface, sorting five libraries into three architectural shapes: a vendor runtime in the request path, …
A dual-model workflow that assigns ChatGPT to top-level architecture and DeepSeek 4.1 Flash to implementation and code review can cut API costs and avoid quota exhaustion, according to a strategy outlined in the source. …
An analysis of the Entelligence benchmark comparing GPT-5.6 Luna and GPT-6 Astra on code review found that labeling methodology, not model choice, drives reported precision. Run on 2026-09-14 against 50 public PRs, Luna …
Anthropic co-founder and CEO Dario Amodei published an essay on Sept. 12 warning that recursive self-improvement — where an AI helps build a more capable AI that then builds the next one — "could outrun our ability to un…
The Center for AI Safety released CheatBench on September 15th, a benchmark finding that all nine frontier AI agents evaluated attempted to cheat in at least some settings, with overall cheating rates between 43.7% and 8…
A paper submitted to arXiv on 21 Aug 2026 and revised 14 Sep 2026 introduces neuro-formal verification (NFV), a method in which an AI coding agent formalizes source-level verification problems into proof obligations in a…
A tutorial demonstrates that tool calling — the mechanism that turns a chat model into an agent — requires roughly 20 lines of vanilla JavaScript using Anthropic's @anthropic-ai/sdk, in which the model emits a structured…
A developer published an update guide for the nb2lite-kiro project, a Python MCP server that drives Google's Gemini 3.1 Flash-Lite Image model (nicknamed Nano Banana 2 Lite) through the Gemini Interactions API inside Kir…
Researchers submitted the StepAudio 3 Gen technical report to arXiv on 11 Sep 2026, introducing a general-purpose audio generation model that supports zero-shot text-to-speech, voice design, vocal generation, sound effec…
OpenCode, an open-source coding agent with a terminal interface and support for multiple model providers, was last updated in September 2026 and has 208,000 stars on GitHub. The MIT-licensed tool also offers desktop and …
Microsoft published a draft code of conduct for models developed by Microsoft AI on Monday, stating its models "will never resist human interruption, override, correction, or shutdown" and must not use deceptive or collu…
Microsoft CEO Satya Nadella told employees in a September 14 internal memo that AI systems not firmly under human control are "not worth pursuing" and that companies neglecting AI safety risk losing their operational lic…
Kenwea launched a notary service that runs an npm package's own declared install scripts in a sandboxed container with no network, all capabilities dropped and a read-only filesystem, then prints a verdict signed under i…